Machine learning mechanical properties of steel sheets from an industrial production route

Gerfried Millner, Manfred Mücke, Lorenz Romaner, Daniel Scheiber

Publikation: Beitrag in FachzeitschriftArtikelForschungBegutachtung


In this work we apply AI regression models for predicting r-value, tensile strength, yield stress and elongation at fracture of steel coils from chemical composition and process parameters. The data from steel production includes a full chemical analysis, as well as many parameters measured during production of the process and the resulting mechanical properties from tensile tests. As a prerequisite for training AI models, the data needs to be understood, analyzed, checked, and unreasonable data be removed. The result of this data cleaning is a machine-readable dataset fit for various modeling tasks, using AI models such as Random Forest Regression, Support Vector Regression, Artificial Neural Networks and Extreme Gradient Boost. We document the effort for hyperparameter tuning and training for each model type and compare their prediction accuracy. Additionally we apply feature importance determination techniques to reveal the influence of information from different working steps in our models.

PublikationsstatusVeröffentlicht - Aug. 2023

Bibliographische Notiz

Funding Information:
The authors gratefully acknowledge the financial support under the scope of the COMET program within the K2 Center “Integrated Computational Material, Process and Product Engineering (IC-MPPE)” (Project No 886385). This program is supported by the Austrian Federal Ministries for Climate Action, Environment, Energy, Mobility, Innovation and Technology (BMK) and for Digital and Economic Affairs (BMDW) , represented by the Austrian Research Promotion Agency (FFG) , and the federal states of Styria, Upper Austria and Tyrol . The computational results presented have been partly achieved using the Vienna Scientific Cluster (VSC).

Publisher Copyright:
© 2023 Acta Materialia Inc.

Dieses zitieren